Dump all properties to CSV/JSON

[jpmckinney]

• [x] BC (letters, notes, files: url, media_type, byte_size, number_of_pages, number_of_rows, duration)
• [x] NL (download_url, media_type)

Fields:

• [x] byte_size
• [x] number_of_rows
• [x] duration
• [x] applicant_type: enum
• [x] fees_paid: pattern
• [x] media_type
• [x] download_url: format uri

Makes it possible to diff all the data over time (not just the core data).